The present invention relates to an image analysis method, device, system, and program, which use vehicle driving information, and a recording medium, and according to one embodiment of the present invention, an image analysis method of a device linked with a camera and a controller through an in-vehicle communication network can comprise the steps of: receiving vehicle driving information from at least one controller for controlling vehicle driving; receiving image information captured by the camera; generating a first optical flow pattern estimated on the basis of the vehicle driving information; generating a second optical flow pattern calculated on the basis of the image information; and identifying a foreground movement pattern on the basis of the first optical flow pattern and the second optical flow pattern. Therefore, the present invention can more accurately analyze, by using vehicle driving information, an image captured by a camera.
Legal claims defining the scope of protection, as filed with the USPTO.
1. An image analysis method in a device operatively connected to a camera and a controller through an in-vehicle communication network, the image analysis method comprising: receiving vehicle driving information from a controller for controlling driving of a vehicle; receiving image information captured by the camera; generating a first optical flow pattern estimated based on the vehicle driving information; generating a second optical flow pattern calculated based on the image information; and identifying a motion pattern based on the first optical flow pattern and the second optical flow pattern, the motion pattern including a foreground motion pattern and a background motion pattern; identifying a moving object based on the foreground motion pattern included in the motion pattern; and correcting camera setting parameters of the camera based on the background motion pattern by adjusting at least one of an orientation angle of the camera and a direction of the camera, wherein the motion pattern is identified based on a difference value between optical flow vectors of the first optical flow pattern and the second optical flow pattern at corresponding positions thereof, and wherein the optical flow vectors are confirmed as the background motion pattern when the difference value is within a predetermined reference range.
2. The image analysis method according to claim 1 , wherein the vehicle driving information comprises steering angle information and vehicle speed information, and the step of generating the first optical flow pattern comprises: estimating a direction of an optical flow vector based on the steering angle information; estimating a magnitude of the optical flow vector based on the vehicle speed information; estimating the optical flow vector based on the estimated direction and magnitude; and generating the first optical flow pattern according to the estimated optical flow vector.
3. The image analysis method according to claim 1 , wherein the step of generating the second optical flow pattern comprises: extracting image frames from the image information; calculating an optical flow vector using two successive ones of the image frames; and generating the second optical flow pattern according to the calculated optical flow vector.
4. The image analysis method according to claim 1 , wherein, in a case in which the difference value deviates from the predetermined reference range, the optical flow vectors are confirmed as an object change pattern.
5. The image analysis method according to claim 4 , wherein the background motion pattern and the object change pattern are generated based on the confirmation results.
6. The image analysis method according to claim 5 , wherein the moving object is identified based on the object change pattern.
7. An image analysis device comprising: a communication unit for transmitting and receiving vehicle driving information and image information; a camera correction unit for correcting setting parameters of a camera; an optical flow pattern generation unit for generating a first optical flow pattern estimated based on the vehicle driving information and a second optical flow pattern calculated based on the image information; a motion pattern identification unit for identifying a motion pattern based on the first optical flow pattern and the second optical flow pattern, the motion pattern including a foreground motion pattern and a background motion pattern; and an object detection unit for identifying a moving object based on the foreground motion pattern included in the motion pattern, wherein the motion pattern is identified based on a difference value between optical flow vectors of the first optical flow pattern and the second optical flow pattern at corresponding positions thereof, wherein the motion pattern identification unit confirms the optical flow vectors as the background motion pattern when the difference value is within a predetermined reference range, and wherein the setting parameters of the camera are corrected based on the background motion pattern by adjusting at least one of an orientation angle of the camera and a direction of the camera.
8. The image analysis device according to claim 7 , wherein the vehicle driving information comprises steering angle information and vehicle speed information, and the optical flow pattern generation unit comprises: a first optical flow pattern generation unit for generating the first optical flow pattern based on a direction of an optical flow vector estimated based on the steering angle information and a magnitude of the optical flow vector estimated based on the vehicle speed information; and a second optical flow pattern generation unit for extracting image frames from the image information, calculating an optical flow vector using two successive ones of the image frames, and generating the second optical flow pattern according to the calculated optical flow vector.
9. The image analysis device according to claim 8 , wherein in a case in which the difference value deviates from the predetermined reference range, the motion pattern identification unit confirms the optical flow vectors as an object change pattern.
10. The image analysis device according to claim 9 , wherein the moving object is detected based on the confirmed object change pattern.
11. An image analysis system comprising: an image capture device for collecting image information of a vehicle that is being driven using at least one image sensor, provided in the image analysis system; an image capture device correction unit for correcting setting parameters of the image capture device; a communication unit for receiving vehicle driving information and the image information through an in-vehicle communication network; and an image analysis device for generating a first optical flow pattern estimated based on the vehicle driving information and a second optical flow pattern calculated based on the image information and identifying a motion pattern based on the first optical flow pattern and the second optical flow pattern, and identifying a moving object, wherein the motion pattern comprises a background motion pattern and a foreground motion pattern, wherein the moving object is identified based on the foreground pattern included in the motion pattern, and wherein the setting parameters of the image capture device are corrected based on the background motion pattern by adjusting at least one of an orientation angle of the camera and a direction of the camera.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
November 23, 2017
February 2, 2021
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.